Key Vocabulary

The Race tracks one central question:

Is AI concentrating power or dispersing capability?

This page defines the core terms used across the site. It is the quick-reference version of the framework: short, plain-language, and meant to help readers follow the argument without getting lost in jargon.

Concentration

Concentration means AI increases the advantage of already-powerful actors.

That power may sit with dominant technology companies, governments, security agencies, wealthy firms, elite institutions, technical experts, data owners, or compute providers.

Concentration often appears through control over compute, data, platforms, proprietary models, vendor relationships, surveillance capacity, or regulation that protects incumbents.

It is not always intentional. It is often the default path when powerful tools move through existing hierarchies.

Dispersion

Dispersion means AI gives more people and institutions meaningful tools, agency, and decision-making power.

It is not just wider access to chatbots. It means AI capability becomes usable, governable, and beneficial across more of society.

Dispersion may appear through open tools, local or private AI systems, worker-centered deployment, public-sector capacity, transparent procurement, civic participation, education, and accountability mechanisms.

Dispersion is not automatic. It usually has to be designed.

Civic Infrastructure

Civic infrastructure is the machinery democracy needs to work in practice.

It includes the systems, institutions, habits, forums, rules, and feedback loops that allow people to learn, deliberate, act, measure, correct, and hold power accountable.

AI could strengthen civic infrastructure by helping public institutions and communities solve problems. It could weaken it by moving decisions into opaque systems ordinary people cannot understand or contest.

Transition Shock

Transition shock is the disruption that occurs when AI changes work, services, skills, authority, or expectations faster than people and institutions can adapt.

It can show up as job loss, workplace compression, skill devaluation, public confusion, institutional overload, loss of trust, uneven access, elite capture, or policy lag.

Transition shock is not only about employment. It is about the gap between technological change and social adaptation.

State Capacity

State capacity is the ability of public institutions to act competently, lawfully, transparently, and effectively.

In the AI era, governments need enough capacity to understand AI, procure it responsibly, evaluate vendors, protect the public, and build internal expertise.

Weak state capacity creates concentration risk. Strong state capacity can support dispersion, accountability, and better public services.

Democratic Accountability

Democratic accountability means people can understand, question, contest, correct, and influence decisions that affect them.

For AI systems, accountability may require transparency, human review, appeal rights, audits, public reporting, procurement standards, feedback channels, clear responsibility, and limits on use.

A system is not accountable just because a public agency uses it. Accountability has to be built into the deployment.

Public Infrastructure

Public infrastructure is the shared capacity society relies on to function and solve problems.

In the AI era, public infrastructure may include public data systems, public-interest AI tools, civic technology, transparent procurement systems, shared learning infrastructure, and governance capacity.

If AI is treated only as a private product, concentration pressures will be strong. If some AI capability becomes public infrastructure, democratic dispersion becomes more plausible.

Default Deployment Path

The default deployment path is the direction AI tends to move when no deliberate countervailing choices are made.

Right now, that path appears concentration-biased. The strongest incentives favor large firms, capital-rich institutions, proprietary platforms, compute owners, data holders, and expert-managed systems.

That does not mean concentration is inevitable. It means dispersion will require institutions, policy, public investment, civic infrastructure, and democratic pressure.

Policy Levers

Policy levers are the tools governments and public institutions can use to shape AI deployment.

They include procurement, public investment, competition policy, labor standards, education and training, transparency rules, data governance, compute access, public-interest research, local experimentation, accountability mechanisms, and civic participation requirements.

The key question is not simply whether AI is regulated.

The better question is what regulation does to power.

The Core Test

When reading any AI development, ask:

Who gains capacity? Who loses agency? Who gets to decide?

That is the vocabulary of The Race.